Parallel benchmarks (for MPI systems only)
并行基准测试（只针对 MPI系统）
A message passing interface (MPI) manages this buffer.
消息传递接口(MPI)管理此缓冲区。
Support for HPC clusters and MPI, including debugging support
支持HPC集群和MPI，包括调试的支持
Parallel programming using MPI to multiply a matrix by a vector.
使用MPI进行并行编程来实现矩阵与向量的乘法。
The MPI also does a better job of uncovering long-term trends.
MPI还对揭示长期趋势有很好的作用。
New Shared memory MS-MPI implementation for multicore servers.
为多核服务器提供最新的共享内存MS -MPI实现。
The MPI home page provides information on message-passing libraries.
MPI主页提供了关于消息传递库的信息。
To compile MPI for the device type ch_p4mpd, use --with-device=ch_p4mpd.
要为 ch_p4mpd设备类型编译MPI，请使用 --with-device=ch_p4mpd。
In order to uninstall MPI from your system, follow the steps listed below.
要从系统中卸载mpi，请按照下面列出的步骤进行操作。
The MPI code needs to be available in a temporary folder, such as /tmp.
MPI代码需要能在临时文件夹中获得，如 /tmp。
It USES an MPI implementation derived from MPICH that we used in our cluster.
它使用了一个源自于本文中使用的MPICH的MPI实现。
In order to build and install MPI, log in as user root on n1 from the desktop.
为了构建和安装mpi，以root用户的身份从台式计算机登录到n1。
Support for developing MPI-based application including MPI cluster debugger.
对开发基于MPI的应用程序的支持，包括MPI集群调试器。
At this stage, MPI has been successfully installed on your machine — for example, n1.
在此阶段，MPI已成功地安装到了计算机(如n1)上。
Run the make command, which compiles the mpi code and makes it ready for installation.
运行make命令，该命令将编译MPI代码并使之为安装做好准备。
Any program that USES MPI API calls can now utilize the compute power of these machines.
现在，任何使用MPIAPI调用的程序都可以利用这些计算机的计算能力了。
Mpich is a software package that USES MPI and provides for efficient parallel computing.
Mpich使用了MPI，是一个用于高效并行计算的软件包。
The parallel version of the Linpack benchmark requires that an MPI library be installed.
Linpach基准的并行版本要求安装一个MPI库。
We look at parallel programming using MPI and the basics of setting up a Linux cluster.
我们将介绍如何使用MPI进行并行编程，以及构建Linux集群的基础知识。
Update the /etc/bashrc file for the PATH variable in order to pick up the right MPI run
更新 /etc/bashrc文件中的PATH变量，以找到正确的MPI 并进行运行
There you have it. Within the MPI framework, four copies of the same process have been created.
在mpi框架中，已经创建了相同进程的四个拷贝。
This article introduces the IPC and MPI libraries along with some of the functionality they offer.
本文介绍IPC和mpi库以及它们提供的一些功能。
You can repeat the above steps for building and installing MPI on all the machines, such as n2 and n3.
可以重复上面的步骤在所有的计算机(如n2和n3)上构建和安装mpi。
The first version was built primarily to support clustering, job scheduling, and MPI 2 based applications.
HPC Server 2008的第一个版本主要着重于支持集群、任务计划以及基于MPI2的应用程序。
The MPI API has several functions that you can apply directly to solve this problem, as shown in Listing 1.
MPIAPI有多个函数可以直接解决这个问题，如清单1 所示。
Boost MPI is not a stand-alone library in the sense that you download it, build it, and you are ready for work.
BoostMPI不是一个单独的库，不能在下载和构建之后直接使用。
Later in this article, we'll set up an MPICH-based Linux cluster and see an example of an MPI-based program.
在本文稍后的部分中，我们将构建一个基于MPICH的Linux集群，并来看一个基于MPI的程序例子。
As shown in Figure 1, the Message Passing Interface (MPI) library specification passes the data between the nodes.
如图1所示，通过消息传递接口(mpi)库规范在节点间传递数据。
Download hpl.tgz, the parallel (MPI) version of the LINPACK benchmark, from netlib.org (see Resources for a link).
从 netlib.org下载hpl.tgz，这是 LINPACK基准测试的并行（MPI）版本（请参阅参考资料 中的链接）。
You don't need to use sockets or other level communication primitives; the MPI back end manages all the hard work.
不需要使用套接字或其他通信原语；mpi后端管理所有底层处理。
应用推荐